Current Control with Self-Learning Ability for PMSM Drives
Author
Cai, Rui
Term
4. term
Education
Publication year
2022
Pages
33
Abstract
Permanent magnet synchronous motors (PMSMs) are widely used in high-performance drives where field-oriented control (FOC) with an inner current loop and outer speed loop is standard. The current loop directly determines torque, yet conventional PI controllers limit transient speed, and predictive current control is sensitive to parameter variations. This thesis analyzes and designs an adaptive PID controller with self-learning capability for FOC-based PMSM drives, enabling online adjustment to maintain performance as machine parameters change. The approach is compared with a traditional PI current loop and evaluated experimentally on a laboratory setup, including tests under parameter changes. Results indicate faster current response and improved robustness to parameter variations with the adaptive PID, and experience-based tuning guidelines are provided. While detailed datasets are not included in the excerpt, the report concludes that the proposed method is effective and practical.
Permanentmagnet-synkronmotorer (PMSM) bruges bredt i højtydende drev, hvor feltorienteret regulering (FOC) med en indre strøm- og ydre hastighedsløkke er standard. Den indre strømregulator er afgørende for momentet, men klassiske PI-regulatorer begrænser den transiente hastighed, og forudsigende strømstyring er følsom over for parametervariationer. Dette arbejde analyserer og designer en adaptiv PID-regulator med selv-lærende egenskaber til FOC-baserede PMSM-drev, der kan tilpasse sig online og bevare ydeevnen, når maskinparametre ændrer sig. Metoden sammenlignes med en traditionel PI-strømsløjfe og evalueres eksperimentelt på et laboratorieopstillet drev, inklusive tests med ændrede maskinparametre. Resultaterne viser hurtigere strømrespons og forbedret robusthed over for parametervariationer for den adaptive PID, og der gives erfaringsbaserede retningslinjer for tuning. Uddraget giver ikke fulde måledata, men rapporten konkluderer, at den foreslåede metode er effektiv og praktisk anvendelig.
[This apstract has been generated with the help of AI directly from the project full text]
Keywords
PMSM drive ; APID ; self-learning ; FOC